Author: IRPA AI Analyst & Senior Advisor, Chris Surdak
In a report released in January 2025 The European AI Action Summit published its International AI Safety Report outlining the concerns, efforts and recommendations for AI governance of 96 AI experts from around the world. Similar to dozens of such efforts around the world, this summit consisted of individuals who recognize that all technologies bring both benefits and costs, and the greater the potential for the former, the greater the potential for the latter.
As the world struggles to absorb the full impact of announcements such as China’s DeepSeek D1, and thereafter OpenAI’s O3 models, such experts continue to work to keep a sense of grounding and risk awareness as trillions of dollars of investment flows into this global effort to dominate AI. As the capabilities and hence adoption of these platforms expand, so too does the plethora of risks that they introduce into our society and economy.
Rather then appearing as luddites, dramatists or Chicken-Little’s, collections of experts such this should be viewed as “the adults in the room,” providing a clear-headed view of the inherent down sides of any new technology. Innovation is great, but the greater the potential leap a technology provides the greater the potential fall that attends the benefit. Wile E Coyote always believes his rocket skis will get him across the desert canyon, but rarely if ever does he complete the jump.
This view does not implore us to chase new tech, rather, it implores us to be aware of, acknowledge and anticipate the potential hazards they hang out with. This viewpoint doesn’t necessarily push for a stoppage or even slow-down in our AI efforts, but rather to maintain a concerted parallel effort to understand and manage the implications of their fruits.
While it may seem contradictory, race cars have brakes so that they can travel faster. The ability of a car to circle a race track quickly is enabled by powerful brakes; tools that allow a driver to quickly get themselves out of trouble when it occurs. Similarly, AI governance is and must be viewed as an enabler of AI adoption, if only to give users a sense that they can get themselves out of any trouble the technology might lead them into.
The team at this conference posed five key questions that practitioners must keep in mind as they drive towards ever-more-powerful models:
How rapidly will general-purpose AI capabilities advance in the coming years, and how can researchers reliably measure that progress?
- What are sensible risk thresholds to trigger mitigations?
- How can policymakers best gain access to information about general-purpose AI that is relevant to public safety?
- How can researchers, technology companies, and governments reliably assess the risks of general-purpose AI development and deployment?
- How do general-purpose AI models work internally?
- How can general-purpose AI be designed to behave reliably?
While these questions were posed at a society-wide perspective, they are fractal; they apply equally to organizations and even individuals. If you personally use these tools as part of your daily life how do you deal with any errors they make? How can you confirm that they are giving you accurate, unbiased information, and how do you respond if you discover they were incorrect? As many attorneys around the world have discovered to their chagrin, you can never really be sure that what an AI tells you is true or false.
A key perspective to maintain is that AI governance needs a seat at the table of those driving the car forward. AI governance is not an add-on. It cannot be viewed as a cost of doing business, otherwise that is all it will be. If, instead, it is viewed as the brakes on the race car, enabling faster adoption with reduced risk, it is likely to produce additional value and return on investment.
As the 298 page report by this commission clearly outlines, today is the day to choose wisely, as the AI investment car accelerates ever-faster towards a transcendental finish line.
About the Author
Chris Surdak is a Senior IRPA AI Advisor and was formerly White House Chief Transformation officer, Automation & AI Practice Lead at EY & Executive Partner for Digital Transformation at Gartner. He’s an engineer, futurist, transformation executive and best-selling author, with over 30 years’ experience in technology development and deployment, digital transformation, blockchain, data and analytics and AI & intelligent automation.
CLICK HERE TO SCHEDULE AN ANALYST CHAT
Links:
Originally posted on 2025-02-05 in the IRPA AI Network — Enterprise AI